When visual information is acquired using a detector mounted to a prosthetic device worn by a patient, the information can include artifacts that reflect involuntary movement of the patient rather than intentional positioning of the detector. These artifacts can limit the usefulness of the acquired visual information. Methods for processing acquired images using various software algorithms in an attempt to reduce such artifacts have been described. As the amount of visual information acquired increases, the hardware requirements for implementing such algorithms to produce relatively rapid results also increase.